I'm too dumb for Zig's new IO interface
openmymind.net
openmymind.net
I finally got it working. I had to flush both the encrypted writer and then the stream writer. There was also some issues with reading. Streaming works, but it'll always return 0 on the first read because Writer.Fixed doesn't implement sendFile, and thus after the first call, it internally switches from streaming mode to reading mode (1) and then things magically work.
Currently trying to get compression re-enabled in my websocket library.
(1) https://github.com/ziglang/zig/blob/47a2f2ddae9cc47ff6df7a71...
Here's an excerpt from the close(2) syscall description:
RETURN VALUE close() returns zero on success. On error, -1 is returned, and errno is set to indicate the error.
ERRORS EBADF fd isn't a valid open file descriptor.
EINTR The close() call was interrupted by a signal; see signal(7).
EIO An I/O error occurred.
ENOSPC
EDQUOT On NFS, these errors are not normally reported against the first write which exceeds the available storage space, but instead against a subsequent
write(2), fsync(2), or close().
See NOTES for a discussion of why close() should not be retried after an error.
And, the moment you start flushing correctly: if(flush(...)) { abort(); }, it becomes infallible from the program's point of view, and can be safely invoked in destructors.
File closure operations, on the other hand, do have legitimate reasons to fail. In one of my previous adventures, we were asking the operator to put the archival tape back, and then re-issuing the close() syscall, with the driver checking that the tape is inserted and passing the control to the mechanical arm for further positioning of the tape, all of that in the drivers running in the kernel space. The program actually had to retry close() syscalls, and kept asking the operator to handle the tape (there were multiple scenarios for the operator how to proceed).
Why can't the OS itself do the prompting in this case, as part of processing the original close()? MS-DOG had its (A)bort/(R)etry/(I)gnore prompt for failing I/O operations, and AmigaOS could track media labels and ask the user to "insert $MEDIA_LABEL in drive".
mov si, GREETINGS_STRING
print_loop:
lodsb ; Load next byte into AL, advance SI
cmp al, 0 ; Check for null terminator
je done
mov ah, 0Eh ; BIOS teletype output
mov bh, 0 ; Page number = 0
mov bl, 07h ; Light gray on black in text mode
int 10h ; Print character in AL
jmp print_loop
done:
...
GREETINGS_STRING db "Hello, BIOS world!", 0
Whatever facility it uses for showing kernel panics, perhaps. Though one could also use IPC facilities such as dbus to issue a prompt in the session of whatever user is currently managing that media device.
Retrying close() is dangerous, if the file descriptor was successfully deallocated, it might have already been re-allocated by another thread. I'd guess the program you're describing was single threaded though (it can still bite there though)
How could one fix that though? It seems pretty unavoidable to me because write() is more or less asynchronous to actual disk I/O.
You could add finalize() which is distinct from close(), but IMHO that's even more confusing.
If the tradeoff was absolute performance/avoiding introducing load-bearing performance-lowering abstraction I think that goal was achieved, but DX may have gone out the window.
Not trying to imply that’s an explicit goal (probably instead just a resource problem), but an observation
Generally speaking I think it is the right trade off for now. Purely inferring from Andrew and the Zig's team online character as I don't know them in person, I think they do care a lot of DX, things like compiling speed and tools. So I think once 1.0 come I won't be surprised if it will have extremely good documentation as well.
And I would argue, writing good, simple, clear, detailed documentation is actually harder than writing code itself.
In 2018, seven years ago, Andrew announced he'd go full-time on Zig and quit his paying job to live off donations instead.
In 2020, so five years ago, Zig's 501(c)3 the ZSF was announced, to create a formal structure to hire more people in addition to the few already on Zig.
So, "most of its time" is just not true. For "most of its time" Zig was a small, largely independently funded project for multiple people, for a tiny period it was a part-time project, and for a while after that it was solo, but those weren't the majority of its existence.
I like to compare this to real world cathedral building. There are some cathedrals that are literally taking centuries to build! It's OK if the important, but difficult thing takes a long time to build.
If you keep up the development pace you're going to approach stability. Unless you're in a manic spiral of rewrites.
Making a programming language from scratch is a long endeavor when it's a one man project.
They unarguably cause confusion for everyone as they change.
But it lets you choose the right abstractions that are going to stick for decades.
If you're going to make a python2 -> python3 transition in your language, make sure it's X0 -> X1.
That said, others have pointed out that writing documentation and tests helps improve quality quite a bit, and in this case it would also increase usability. I think I'd agree with this stance, but there is no way I could make the statement that even most of the code I've written for public consumption had excellent documentation or examples. So I've got no leg to stand on there, just the armchair.
> And I would argue, writing good, simple, clear, detailed documentation is actually harder than writing code itself.
All the more reason why it must be done! A little silly but from my armchair maybe it's one of those "start with the interface you want and work backwards", but the problem is that approach can be at odds with mechanical sympathy and we know which side Zig lands on (and arguably should land on based on it's values).
If so, I believe Zig will stay within a niche. Lower entry barriers allow "script kiddies" to easily start withe language, and they eventually will become leading engineers. Only a few people tend to go straight for the highest practice without "playing around". IMHO the reason, why PHP got so popular (it was not good back then, just very very easy to start with).
Yes.
I think a contributor that really wanted to help the ecosystem would start in the stdlib and then start moving outwards. Even if it was LLM-assisted, I think it could be high value.
IIRC Loris already has an engine for building websites with Zig, but making sure that every Zig library has docs (similar to rustdocs) might be a great start. It is incredibly useful to have a resource like rustdocs, both the tooling and the web sites that are easily browsable.
Again, maybe everyone in the Zig ecosystem just has amazing editor setups and massive brains, but I personally really like the ease of browsing rustdoc.
> If so, I believe Zig will stay within a niche. Lower entry barriers allow "script kiddies" to easily start withe language, and they eventually will become leading engineers. Only a few people tend to go straight for the highest practice without "playing around". IMHO the reason, why PHP got so popular (it was not good back then, just very very easy to start with).
I agree, but I'd add that the niche they're aiming for is systems programming, so they're probably fine :). The average hacker there is expecting C/C++ or to be near the metal, and I think Zig is a great fit there. They're likely not going to convince people who write Ruby, but it feels reasonable for C hackers.
Also I want to just be clear that I think Zig has a lot of motivating factors! They're doing amazing things like zig cc, unbelievably easy, "can't believe it's not butter" cross-compilation, their new explicit/managed I/O mechanism, explicit allocators as a default, comptime, better type ergonomics. It's a pretty impressive language.
Tbh, this sort of auto-generated docs from source code is not all that useful, since you get that same information right in the IDE via the language server.
The important documentation part that's currently missing is how everything is supposed to work together in the stdlib, not the 'micro-documentation' of what a single type or function does. And for this sort of information it's currently indeed better to look at example code (e.g. the stdlib's testing code).
IMHO it's way too early for this type of high-level documentation, since things change all the time in the stdlib. Putting much work into documenting concepts that are discarded again anyway doesn't make much sense.
I am just editing docs now that Claude Code writes for me. I am fanatic about developer docs (and I guess an exception as I love writing them) but with a set of concise instructions for CC and some writing style examples I get 90% there, sometimes 99%.
If you believe you don't have time for the last 1--10% you should not be in charge of writing any API used by anyone but yourself. Just my two c.
Lack of docs also cripple AI from understanding, so future adoption becomes even more bleak.
If an api or library developer didnt bother doing even bare minimum docs, my confidence in the library drops aswell.
Did they skip testing aswell? Ran the happy path for a day and called it good?
This post sour my interest in zig. Its now obvious to me now why rust took much of its market.
Zig (programming language) - First appeared 8 February 2016; 9 years ago
Rust (programming language) - First appeared January 19, 2012; 13 years ago
Also, Zig at this point isn't really a brand new language anymore. I have comments on their issues dating back to 2018, so it's been a very active language since at least then.
Just getting started is an even bigger reason to have good docs to clearly communicate how the libraries and APIs work!
I wouldn't even read a push request containing a new function if the creator didn't bother writing a short description and usage clarification.
Getting started is a good excuse for limited libraries or support (same situationwith rust). But lack of even basic docs is not acceptable if you want user adoption.
Of course, this isn't meant to be a defense for the lack of documentation on Zig's side, but in my experience, Zig's code definetly is much easier to read, just because for the fact, that Rust's std code is akin to C++'s stl.
One of the personal grimes i have with Zig is, that `anytype` makes the function contract kind of meaningless, because you can't see what is expected purely on the function definition.
I think what I’ve come to realize is that when I feel a barrier toward doing that work, it’s a sign that I don’t actually like the underlying API. I don’t want to document it or craft examples or tutorials because in my mind the API sucks, it’s transitional, and documenting it in its current state is going to somehow lock in a bad interface and further increase the effort required to change and fix it up later.
I tried Zig a couple of times and I got that feeling: very powerful and clever language but not really for me, I don't have the headspace, sorry. I need something I can debug after an 8 hours dayjob, a commute and having put the kids to bed. It better be inviting & fun! (Hi, C).
(Loris Cro being a key community figure isn't helping in any way, and it's a good remainder that if you don't clear up your community from bullies from the beginning, they will turn your entire community to a miserable place. And that's a shame because from what I've seen, Andrew Kelley seems to be a very cool guy in addition to being very smart).
It is the anti-intelectualism from Go culture, gone wild against C++, Rust, Swift, anything modern, or even tools, using game engines versus doing the whole computer from scratch for a game.
Andrew's talk is here (second event after the two people chatting while sitting on chairs): https://handmadecities.com/media/seattle-2024/hms-day-one/
Here you can see a particularly funny (but also sad) reaction by one of these people https://drive.proton.me/urls/MB1EB4EF34#YZdvmAvBFp1C
> using game engines versus doing the whole computer from scratch for a game
That said you are doing yourself a disservice if you think that not using an engine to make a game is a form of "anti-intellectualism".
Depends on the attitude, not using an engine, because one wants to learn the whole stack, makes all sense, after all people need to learn how to make game engines, if nothing else for the next generation.
Not using one out of spite, because we do everything handmade over here attitude, is a completely different matter.
https://wiki.xxiivv.com/site/2025.html (the entry under 19b)
It seems this was a right vs. left (or liberal) split.
All this, combined with the fact that Zig, at best is still in beta quality and at worst amounts to a massive waste of everyone’s time, makes it unsurprising that people block you and simply refuse to engage with your loud community efforts, endless churn and crust tied to beta quality compiler.
> Zig is not really a handmade project, case in point both Andrew and I are blocked on social media by the two gods of the handmade movement (casey and john) and, according to their die hard fans, Andrew gave a talk at the last handmade conference that caused the community to split apart (the reality is a bit more complex than this, but Andrew's talk is certainly one that you wouldn't see at their new "better software" conference).
> Andrew's talk is here (second event after the two people chatting while sitting on chairs): https://handmadecities.com/media/seattle-2024/hms-day-one/
> Here you can see a particularly funny (but also sad) reaction by one of these people https://drive.proton.me/urls/MB1EB4EF34#YZdvmAvBFp1C
Regarding the links you posted:
In the first, at 2:30:40, Andrew Kelly publicly calls out a specific author of a competing technology in exaggerated, caricatured, and fabricated context.
In the second video, yet another author of a yet another competing technology directly points out this unapologetic and concerning behavior on Andrew Kelly’s part.
And now you—“VP of Community @ Zig Software Foundation”—assert your “righteous” stance by sharing these videos, while ironically pointing out that some of those same individuals (of competing technologies fame) block you on social media.
Too bad that doing your job probably means being as loud and visible online as possible to spread the molecules of Zig no matter what.
Maybe, or maybe the fact that Zig is a small independent project with limited resources has also something to do with it, and this kind of shaming says less about Zig than you'd think.
When I first joined the Zig project, Zig was still using the bootstrap compiler written in C++ that would not free memory (it took more than 4GB to compile the Zig compiler). Some people at the time were asking us to prioritize work on the package manager but Andrew rightfully wanted to prioritize rewriting the compiler instead. In hindsight this was the obviously right decision: a package manager implies that one can very easily add an order of magnitude more code to their project, stressing the performance of the compiler. If we had not prioritized core infrastructure over giving people what they wanted faster, today we would have people complaining that adding a single dependency to their project makes the build impossible to complete.
The Zig project has a huge scope and we are a small independent organization. This makes us extremely nimble and efficient, but it does mean that we need to do things in the order that makes the most sense for the project, not for what the public wants.
The fact that we develop in the open doesn't mean that the language is ready yet.
People that already have the required domain knowledge (and who have a tolerance for breaking changes) will have the opportunity to be early adopters if they wish to do so, others will have to wait for Zig to become more mature. And we do make this clear in releases and all forms of public communication.
We have gone a long way since the bootstrap compiler days, but we are still missing key infrastructure:
- we have a x86_64 custom backend but aarch64 is not complete yet - incremental compilation is showing that we can get instant rebuilds of large projects, but it has missing features and it doesn't work on all platforms yet - we need native fuzzing since AFL keeps regressing everytime a new version of LLVM comes out - for the longest time we haven't had a strong I/O story, now we're finally working on it
The time for paving the road for a new generation of programmers will come (it's in the ZSF mission statement btw), but first we need to finish the plumbing.
That sounds strange. Modern C++ requires very little manual memory management, at least when you're writing something high-level like a compiler. C++11 had been out for years when development on Zig started. Were they writing C++ old-school as C-with-classes and malloc() everywhere? Why not use a more appropriate language for the first prototype of a compiler for a brand new language?
Why would you care about these kinds of micro-optimizations at that stage of development, when you don't even know what exactly you need to build? We're not talking about serious algorithmic improvements like turning O(n²) into O(n) here.
> Freeing memory can actually cost you performance, so why not just let the OS clean up for you at exit(2)?
Because a compiler is not some simple CLI tool with a fixed upper bound on resource consumption.
it turns out that compiler speed is bound by a bunch of things, and it's death by a thousand cuts. If you have a slow compiler, and it takes forever to compile your compiler, your language becomes scelerotic, no one wants to make changes, and your language gets stuck in shitty choices.
> Because a compiler is not some simple CLI tool with a fixed upper bound on resource consumption
yes. thats right. a compiler is complex and should use several different allocation strategies for different parts. if your language steers you towards using malloc for everything then your compiler (assuming it's bootstrapped) will suffer, because sometimes there are better choices than malloc.
> When I first joined the Zig project, Zig was still using the bootstrap compiler written in C++ that would not free memory (it took more than 4GB to compile the Zig compiler). Some people at the time were asking us to prioritize work on the package manager but Andrew rightfully wanted to prioritize rewriting the compiler instead. In hindsight this was the obviously right decision: a package manager implies that one can very easily add an order of magnitude more code to their project, stressing the performance of the compiler. If we had not prioritized core infrastructure over giving people what they wanted faster, today we would have people complaining that adding a single dependency to their project makes the build impossible to complete.
I think you may have missed that the intention was always to rewrite the compiler to become self hosted. Improving the C++ implementation any more would've been pointless optimization.
Or, maybe it's this kind of redirection and evidence of a victim complex. Part of the reason that there's a patina of anti-Rust sentiment includes the dismissive attitude and swipes you, a the VP of Community at the Zig Software Foundation, take towards Rust and Rust developers by writing about topics you aren't involved in and don't have a solid grasp of.
https://kristoff.it/blog/raii-rust-linux/ https://lobste.rs/s/hxerht/raii_rust_linux_drama#c_gbn1q8
Or similarly, comments like this one in this thread: https://news.ycombinator.com/context?id=44994749
Loris is a bully, that's the problem, not his opinions.
* Sorry you feel that way, we try not to foster an inclusive environment that avoids language flamewars and negativity towards other languages
* Some specific examples of policies that are enacted in community spaces towards that end (e.g. some large programming language discord servers have specific rules against low effort language bashing content and memes)
* Take the opportunity to link back to positive things that the Zig team has said about Rust and places where folks have recommended it.
pcwalton infamously declared zig was "a massive step back for the industry" https://x.com/pcwalton/status/1568306598795431936?s=46&t=OCi.... He and the Rust Core Team had a big reputation for burning bridges. Even to this day, the new Rust leaders are happy to attack other memory safe languages like Go, declaring them "not memory safe" https://news.ycombinator.com/item?id=4467200
I think Kristoff remembers these attacks, and crucially how very few voices within the Rust community push back against Rust supremacism.
> [Golang] "not memory safe"
Both of these are entirely fair assessments, not "attacks". Golang really does have memory safety issues with concurrent code, and a memory-unsafe language like Zig is a step back even compared to Java/C#, let alone Rust.
The end result is that Rust's leaders either avoid interacting with other languages, or engage in flamewars. I think it's a big reason why Java, the most popular and successful memory safe language in the world, has little-to-no formal contacts with the Rust team.
I can't believe you really wrote that.
> Then [Anti Rust person] should have been perma-banned long ago [on an open forum]. Until this is done, we'll have to warn people about engaging with him…
You can hate kristoff, ignore him or attack his arguments. You can also love a piece of software and treat it as sacred. But other people should not be subjected to that love nor should they be canceled on account of it. Flag it or downvote. Beyond that, it's outside of our control.
I do agree though that kristoff should focus on Zig and not indulge in provoking old enemies. His valid points— deferring premature documentation for newbies until concepts are ironed out— are being lost in programming language holy wars.
I'm no language though…
> You can hate kristoff, ignore him or attack his arguments. You can also love a piece of software and treat it as sacred.
It's not about “love” or “being sacred” or even about Rust or Zig, it is about behaving in society. Most successful communities at some point meet toxic people who want to start holly wars and insult people. Successful communities are the one who ban those people, or at least coerce them into behaving through social pressure.
When you don't do that, you end up with bullies like Loris occupying prominent positions, and that's very bad for the community because it attracts people like that.
It has nothing to do with programming languages at all.
And sorry to say that bluntly, your discourse about “supremacism” “love” or “sacred” sounds very immature: programming languages are tools and engineering projects, not icons that should be worshipped or hated, they all have their strong points and warts (and god knows Rust has its share of annoyances…). Don't get dragged in holy wars by cult leaders like Loris.
Maybe you can make a hierarchy between technical criticism like this and the fact that Go isn't technically memory-safe[1], with Loris' abusive behavior of calling Rust maintainers names like “wankers”…
[1]: which is a criticism mostly coming from the Java crowd, by the way, not Rust, like the criticism of the simplistic garbage collection management in Go
What might be the better critique about this, is that any programming language's leadership should not be engaging in that kind of bad behavior. And any ill words coming from them about another language, should always be taken with a grain of salt and seen as likely bias.
It reflects poorly on any leadership engaging in that kind of bad and unprofessional behavior, and it eventually backfires on any project or person. People eventually notice it and figure out the foul things that have been or is being done, then demand accountability or walk away from the toxicity.
I will hopefully wait for comments from Ghostty, Bun or Tigerbeetle Devs.
On another point that is wroth mentioning, I hope Andrew will at least put it out publicly, IMO Zig isn't Anti-Rust. But it did attract the type of people who are not too happy with Rust. I dont remember a single time Zig came out to bash anything about Rust. It isn't Anti anything at all. Its goal is to be a decent C replacement as very high level assembly languages, aiming at Data Oriented Design with some interesting features, and extremely fast compilation speed. ( Which was the reason why I was interested in it in the first place. I miss Turbo Pascal )
Zig reminds me of the old school, traditional projects. It isn't perfect for everything, it never claims to be, if you like it, use it. If not, there are plenty of options out there.
At least Ghostty Dev seems to be enjoying it almost every day.
But as you say, there's no reason why Zig ought to be anti-Rust, both language are are fresh attempts at low level programming, both highly opinionated and with very different philosophies and trade offs and both language can cohabit peacefully (I've heard good things about using the Zig toolchain for cross compilation of C dependencies in rust projects, so the existence of Zig already has had a positive impact on the Rust ecosystem).
While he can get petty when talking about memory safety, he was never obsessed by Zig in any way, his tweet that ended up with Loris calling him “Coomer” among other names was a response to someone mentioning that Zig wasn't memory safe. I've never seen him mentioning Zig first on Twitter.
And on the other hand you have Loris, obsessed so much by Rust he can't help spawning in Rust threads or write dumb anti-rust rants on his blog…
(In fairness, I must say that Loris isn't even the worst offender, that title goes to M. Presler, who keeps spawning every Rust thread telling everyone who want to hear it that Zig is much better, despite having seemingly no stakes in the game (being a Java guy and having no public Zig contribution or even documented use of the langage). Sounds like the “Rust evangelical strike force” emotionally hurt a bunch of people so much they became lifetime haters).
I have not seen much of any anti-Rust sentiment in the community. There's a lot of people in the community who do Rust, like rust, and work on rust projects. If the Zig community has an anti-anything sentiment, it's against C++.
(You won't have to seek long in his HN history to find an instance of such behavior)
What's not very good are the people who don't like Rust, who are uneasy with Rust is eating the system programming world, and are now pushing Zig as the champion of the resistance about Rust.
It happens a lot, unfortunately.
I didn't.
While tests aren’t quite as good documentation as actual documentation, they are guaranteed to not be out of date.
Something - anything. As much as I like Zig, I dread returning to it after a few months of being out of the loop.
Other options include but are not limited to providing minimal, low effort examples, high-level overview, linking to projects using these features, linking to relevant tests, commits, or source code, setting up an official community wiki and encouraging people to contribute.
Where does a beginner go to learn how to use the package manager these days? It looks like they still won't find any clues in the "Learn" section of Zig's website.
There's a promising page titled "Zig Build System" in there which references a "Package Management" section, but when you click on it, it doesn't exist!
But to answer your question, it exists in the comments of the auto-generated build.zig.zon file
The official answer to complaints about missing documentation has always been "ask in Discord". Pretending that this isn't the case is just disingenuous.
> But to answer your question, it exists in the comments of the auto-generated build.zig.zon file
The comments document the file format of build.zig.zon, they don't tell you anything about how to actually use a dependency in your build system.
Pretty funny, coming from someone who went from "false dichotomy" to "the false equivalency".
You didn't accuse anyone of it, you presented one (https://news.ycombinator.com/item?id=44995714) and then claimed it wasn't one (https://news.ycombinator.com/item?id=44994200).
And then you blithely went on to talk about "the false equivalency" (https://news.ycombinator.com/item?id=44996188), which nobody had been talking about. What's that, if not "moving goalposts" (https://news.ycombinator.com/item?id=44998340)?
And I think if you won't be responding further it's because I either confused you or, more probably, just simply made you realise you were wrong. (Or maybe you knew that all along and just brazenly projected your own goalpost-moving onto someone else, Idunno.)
But if you change your mind and do respond, I'd be grateful for an explanation of how you came to the conclusion that I "seem confused", and about what exactly.
The zig users on this thread seem to not understand this, and all seem to think documentation is a thing you write later for users when everything settles down. Or is somehow otherwise "in the way" of feature and API development speed.
That is a very strange view.
If writing developer documentation is having serious affect on your language feature velocity, you are doing something very wrong. Instead, writing it should make things move faster, becuase, at a minimum, others understand what you are trying to do, how it is going to work, and can help. Also, it helps you think through it yourself and whether what you are writing makes any sense, etc.
Yes there are people who can do this all without documentation, but there are 100x as many who can't, but will still give high quality contributions that will move you along faster if you enable them to help you. Throwing out the ability to have these folks help you is, at a minimum, self-defeating.
I learned this the hard way, because i'm one of those folks who can just stare at random undocumented messy code and know what it actually does, what the author was probably trying to do, etc, and it took years till i learned most people were not like this.
Of course documentation is good. But if you have to prioritize either a new feature, or a critical bugfix, or documentation, you often can't have it all
I do very much prefer moving fast though, so I get it, docs-later is obviously a very valid way of doing things.
If someone is excited about Zig and wanted to make a difference I guess it’s now obvious where they could have outsized impact!
But when working in open-source and your goal is to have people adopt your software, then it's a bad point and a lazy excuse.
There's a reason for the 0.x version number, if you can't live with breaking changes, don't use Zig yet. It's as simple as that.
They target embedded c market. Limited support and features is expected and common. But being lax with docs and testing is not acceptable.
First: Those two things are not the same. The Zig stdlib and compiler have tons of tests, and it's actually quite rare to stumble over implementation bugs.
Second: I'd rather take no docs at all than per-function or per-type doc-headers which just repeat what the code does in natural language (and then may quickly get stale). Some parent wrote about generating docs with an LLM which is just a laughably bad idea, since even with perfect results you'll just get a redundant reiteration of what the code below does, just in imprecise human language, it would literally just be redundant noise.
Documentation mostly makes sense up on the systems level to explain high level concepts, not as 'this function does this and that...' level, this information already exists in much more precise form in the source code below the doc header.
But this sort of high level documentation doesn't make sense as long as a project is in the experimentation phase, and Zig is still one big experiment.
This post itself is a great example. The code is not self explanatory unless you understand the library below. So if this was library code; you would have to reverse through two abstraktion walls to understand it.
And as for high level documentation... it would be very helpful to have that DURING the development process to discuss the implementation, and quickly get people up-to-speed in the plan during code review. Id write that before the code.
I mention tests, because they are 10x more work to write than docs. So if the developer skips out on docs, my trust in their testing is also destroyed.
Are you saying one should never make anything half finished available to the public? This post proves why it is valuable to do so, they are getting valuable feedback and a discussion on hacker news for free.
If the project invites me to use it "for maintaining robust, optimal and reusable software." without putting "super unstable, we don't even care about docs" on the same page... that's also saying something.
if you have that attitude about docs, then likely you are being gatekept from the project until it hits 1.0.
Similar to how TDD forces you to first figure out the API of your code due to the test code being its first client.
If you don't through that exercise, you're much more likely to build confusing, difficult-to-use APIs.
have not found this to be the case with zig in general. you could easily make the opposite argument, that documenting things (especially quirks) can give you license to build confusing APIs.
> have not found this to be the case with zig in general.
Dunno how general it is, but TFA we're discussing here contains an example, so AFAICS it seems at least a little too common.
Code is rarely written from start to finish in a single session. Would you rather spend 5 minutes before you do a git commit to write down some very basic documentation, or spend an hour rediscovering it when you pick up development two weeks later?
Nobody is expecting extensive and well-polished documentation here, but is a "Flags X, Y, and Z are mandatory, buffer A is required: I tested with 10MB, smaller might be fine too" really too much to ask for?
Is the time you saved by not writing that one line worth someone else spending several hours figuring out how to do a "hello world"?
A different answer to that question, could be that its still unstable and making breaking changes (not 1.0), thus a reluctancy to focus on documentation to a level many consider satisfactory.
I would say just stay away from the standard library for now and use your OS API, unless you're willing to be a beta tester.
Personally, I think it is wrong to inflict your experiments on other people and when you pull the rug out from underneath say, well, we told you it was unstable, you should't have depended on us in the first place.
I don't even understand what zig is supposed to be. Matklad seems to think it is a machine level language: https://lobste.rs/s/ntruuu/lobsters_interview_with_matklad. This contrasts with the official language landing page: Zig is a general-purpose programming language and toolchain for maintaining robust, optimal and reusable software. These two definitions are mutually incompatible. Moreover, zig is clearly not a general purpose language because there are plenty of programming problems where manual memory management is neither needed nor desirable.
All of this confusion is manifest in zig's instability and bloated standard library. Indeed a huge standard library is incompatible with the claims of simplicity and generality they frequently make. Async is not a feature that can be implemented universally without adding overhead and indirection because of the fundamental differences in capabilities exposed by the various platforms. Again, they are promising a silver bullet even though their prior attempt, in which they publicly proclaimed function coloring to be solved, has been abandoned. Why would we trust them to get it right a second time?
There are a very small number of assembly primitives that every platform provides that are necessary to implement a compiler. Load/store/mov/inc/jeq/jump and perhaps a few others. Luajit implements its parser in pure assembly and I am not aware of an important platform that luajit runs on that zig goes. I do the vast majority of my programming in lua and _never_ run into bugs in the interpreter. I truly cannot think of a single problem that I think zig would solve better than luajit. Even if that did exist, I could embed the zig code in my lua file and use lua to drive the zig compiler and then call into the specialized code using the lua ffi. But the vast majority of code does not need to be optimized to the level of machine code where it is worth putting up with all of the other headaches that adopting zig will create.
The hype around zig is truly reaching llm levels of disconnection from reality. Again, to believe in zig, one has to believe it will magically develop capacities that it does not presently have and for which there is no plan to actually execute besides vague plans of just wait.
The weird interface of go is probably due the fact that some interfaces can be used to extemd the writer like the hijacker interface (ResponseWriter.(http.Hijacker)) and the request object is used multiple times with different middlewares interacting with it. In short: request does not need to be extended, but the response can be an websocket, an wrapped tcp connection or something else.
You can lift/unlift in or out of arbitrary IO, in some languages one direction is called a mock, in other languages the opposite is called unsafeFoo.
Andrew Kelley independently rediscovered on a live stream 30 years of the best minds in Haskell writing papers.
So the future is Zig. He got there first.
Compute is getting tight, lots of trends, the age of C++ is winding down gracefully. The age of Zig is emerging delibetately, and the stuff in the middle will end up in the same historical trash bin as everything else in the Altman Era: the misfortunes of losing sight of the technology.
The age of C++ is going great, despite all its warts and unsafety, thanks to compiler frameworks like GCC and LLVM, games industry, GPGPU and Khronos APIs.
Even if C++ loses everywhere else, it has enough industry mindshare to keep being relevant.
Same applies to C, in the context of UNIX clones, POSIX, Khronos, embedded.
Being like Modula-2 or Object Pascal in safety, in C like syntax, isn't enough.
Rust makes false promises in practical situations. It invented a notion of safety that is neither well posed, nor particularly useful, nor compatible with ergonomic and efficient computing.
It's speciality is marketing and we already know the bounding box on its impact or relevance. "Vibe coding" will be a more colorful and better remembered mile marker of this lousy decade in computers than Rust, which will be an obscurity in an appendix in 100 years.
"Makes predictions to within a quantifiable epsilon"? What in the world do you mean? The industry experience with C++ is that it is extremely difficult (i.e., expensive) to get right, and C++20 or newer does not change anything about that. Whatever "epsilon" you are talking about here surely has to be very large for a number bearing that sobriquet.
As for the mindless anti-Rust slander... I'm not sure it's worth addressing, because it reflects a complete lack of the faintest idea about what it actually does, or what problem it solves. Let me just say there's a reason the Rust community is rife with highly competent C++ refugees.
I doubt it.
I'm teaching a course on C this fall. As textbook I've chosen "Modern C" by Jens Gustedt (updated for C23).
I'm asked by students "Why don't you choose K&R like everyone else?"
And while the book is from 1978 (ANSI C edition in 1988), and something I've read joyously more than once, I'm reminded of how decades of C programmers have been doing things "the old way" because that's how they're taught. As a result, the world is made of old C programs.
With this momentum of religiously rewriting things in Rust we see in the last few years (how many other languages have rewritten OpenSSL and the GNU coreutils?), the amount of things we depend on that was incidentally rewritten in Rust grows significantly.
Hopefully people won't be writing Rust in 100 years. Since 100 years ago mathematicians were programming mechanical calculators and analog computers, and today kids are making games. But I bet you a whole lot of infrastructure still runs Rust.
In fact, anything that is convenient to Vibe code in the coming years will drown out other languages by volume. Rust ain't so bad for vibe coding.
There is a place to learn about history of computing, and that is where K&R C book belongs to.
Not only is the old way, this is from the age of dumb C compilers, not taking advantage of all stuff recent standards allow compiler writers to take to next level on optimizations, not always with expected results.
Maybe getting students to understand the ISO C draft is also an interesting exercise.
Please stop. Rust's promise is very simple. You get safety without the tracing GC. It also gives you tools to implement your own safe abstraction on top of unsafe, but you are mostly on your own (miri, asan, and ubsan can still be used).
Neither Rust nor Ada nor Lean nor Haskell can guarantee there are no errors in their implementations.
Similarly, none of the listed languages can even try to show that a bad actor can't write bad code or design bad hardware in a way that maintains their promises. If you need that, you need to invent the Omniscient Oracle, not a program.
I hate this oft repeated Nirvana fallacy. Yes, Rust is offering you a car with seatbelts and airbags. It is not offering a car that guarantees immortality in the event of a universe collapse.
Because it's technically true. The best kind of true!
Sorry, I meant to say the opposite of truth. Neither Rust nor Ada.Spark, which use LLVM as a backend, can prove via that they are correct if LLVM has bugs.
In the same way, I can't guarantee tomorrow I won't be killed by a rogue planet hitting Earth at 0.3c. So I should probably start gambling and doing coke, because we might be killed tomorrow.
> Every single project needs to "fix" the same kind of safety issues over and over again
I doubt that's the biggest problem. Each of the unsafe libraries in C/C++/Zig can be perfectly safe given invariants X and Y, respectively. What happens if you have two (or more) libraries with subtly non-compatible invariants? You get non-composable libraries. You end up with the reverse problem of the NPM world.
There are some scary soundness holes in Rust's compiler that will get patched eventually but in principle you could trip them today. They're often "But why would anybody even do that?" problems, but it's technically legal Rust and the compiler doesn't reject your program or even ICE it just miscompiles your input which is not what we want.
Sure. Again, I didn't say there are no bugs in Rust codebase ever, or that Rust will prevent all of errors forever.
They are working on them, but a large chunk ~50% (45 out of 103) are either nightly-only bugs (due to nightly-only features) or LLVM bugs that are difficult to coordinate/solve.
Some part of them will probably be helped or solved by a new trait resolver, Polonius. Granted, this still won't resolve all issues. Or even all trait related issues.
> There are some scary soundness holes in Rust's compiler that will get patched eventually but in principle you could trip them today.
In principle, yes. In principle you should cower in a corner, because there are many scary things in the world. From microbes, to insects, to malevolent proteins, to meteors, to gamma rays, to strangelets, to rogue black holes, to gravity entering a new stable state.
In practice, you don't. Because in practice we empirically adjust our expectations to the phenomena that occur, or have occurred. So a war is very likely, but being vaporized by a stray cosmic ray is not very likely.
And most of those UB require either strange code or a weird combination of hardware, compilers, or flags. It's probably why most people don't encounter UB in everyday Rust.
By that logic, we definitely have enough safe languages as it is, as there are many more. But this safe/unsafe dichotomy is silly, and is coloured by languages that are unsafe in some particular ways.
1. Memory safety is important because memory-safety violations are a common cause of dangerous security vulnerabilities. But once you remove out-of-bounds access, as Zig does, memory safety doesn't even make it to the top 5: https://cwe.mitre.org/top25/archive/2024/2024_cwe_top25.html I.e. the same logic that says we should focus on safety would lead us to conclude we should focus on something else.
2. Memory safety has a cost. To get it, you have to give up something else (there could even be a cost to correctness). That means that you have to consider what you're getting and what you're losing in the context of the domain you're targeting, which is not the same for all languages. C++, with its "zero-cost abstractions", believed it could be everything for everyone. That turned out not to be the case at all, and Zig is a very different language, with different goals, than C++ originally had.
Given Zig's safety guarantees (which are stronger than C++'s), and given its goals (which are different from C++'s), the question should be what should we be willing to give up to gain safety from use-after-free given the language's goals. Would more safety be better if it cost nothing? Of course, but that's not an option. Even Java and Rust could prevent many more dangerous bugs - including those that are higher risk than use-after-free - if they had more facilities like those of ATS or Idris. But they don't because their designers think that the gains wouldn't be worth the cost.
If you don't say what Zig programmers should give up to gain more safety, saying "all new languages should be memory-safe" is about as meaningful as saying we should write fewer bugs. That's a nice sentiment, but how and at what cost?
I am a firm beliver in the vision of Xerox PARC for computing, and think the only reason we aren't yet there are politics, lack of funding from management for doing the right thing pushing them into the market, always looking to shareholders and the next quarter, and naturally programming language religion.
We were already on the right direction with languages like Modula-3 and Active Oberon, following up on Cedar influences, unfortunately that isn't how the industry goes.
Rational started as a company selling Ada Machines, that didn't had such issues with compilation times, but again goes down to reasons I listed why mainstream keeps ignoring such tools until finally governments are stepping in.
What is that in relation to Zig and memory safety? Am I missing some context?
Instead we got UNIX and C.
Had Java actually been designed for where it stands today, instead for set top boxes and browser applets, retrofitting the value types, AOT compilation and low level capabilities already available in Modula-3 and Eiffel wouldn't be such an engineering feat that is already on 15 years mark.
Interestingly enough, value classes approach remind very much how expanded classes work in Eiffel.
Likewise money keeps being burned into making industry adopt a JIT compiler into CPython, which they could have been enjoying with Smalltalk and Lisp.
It may sometimes be the case that we say, oh, that particular feature of some unchosen language can be very useful for us now, but that is something very different from we should have chosen that other language - with all its pros and cons - back then.
Notice that this cost, which proponents of Zig scoff at just like C++ programmers before them, is in fact the price of admission. "OK, we're not correct but..." is actually the end of the conversation. Everybody can already do "Not correct", we had "Not correct" without a program, so all effort expended on a program was wasted unless you're correct. Correctness isn't optional.
A language like Rust exists precisely because correctness isn't the only concern, as most software is already written in languages that make at least as many guarantees. Rust exists because some people decide they don't want to pay the price other languages take in exchange for their guarantees, but they can afford to pay Rust's price. But the very same reasoning applies to Rust itself. If Rust exists because not all tradeoffs are attractive to everyone, then clearly its own tradeoffs are not attractive to everyone.
The goal isn't to write the most correct program; it's to write the most correct program under the project's budget and time constraints. If you can't do it in those constraints, it doesn't matter what guarantees you make, because the program won't exist. If you can meet the constraints, then you need to ask whether the program's correctness, performance, user-friendliness etc. are good enough to serve the software's purpose.
And that's how you learn what software correctness researchers have known for a long time: sometimes increasing correctness guarantees can have unintuitive cost/benefit interactions that they may even end up harming correctness.
There are similar unintuitive results in other disciplines. For example, in software security there's what I call the FP/FN paradox. It's better to have more FN (false negatives, i.e. let some attacks go through) than more FP (false positives, i.e. block interactions that aren't attacks) because FPs are more likely to lead to misconfiguration or even to abandonment of the security mechanism altogether, resulting in weaker security. So, in software security it's a well known thing that to get better security you sometimes need to make fewer guarantees or try less hard to stop all attacks.
In a better, saner world, we'd writing Ada++ not C++. However, we don't live in a perfect world.
> The goal isn't to write the most correct program; it's to write the most correct program under the project's budget and time constraints.
The goal of ANY software engineer worth their salt should be minimizing errors and defects in their end product.
This goal can be reached by learning to write Rust; practice makes perfect.
If GC is acceptable or you need lower compilation times, then yes, go and write your code in C#, Java, or JavaScript.
As someone who worked on safety-critical air-traffic-control software in the nineties, I can tell you that our reasons for shifting to C++ were completely sane. Ada had some correctness advantages compared to C++, but also disadvantages. It had drastically slower build times, which meant we couldn't test the software as frequently, and the language was very complicated that we had to spend more time digging into the minutiae of the language and less time thinking about the algorithm (C++ was simpler back then than it is now). When Java became good enough, we switched to Java.
Build times and language complexity are important for correctness, and because of them, we were able to get better correctness with C++ than with Ada. I'm not saying this is universal and always the case, but the point is that correctness is impacted by many factors, and different projects may find achieving higher correctness in different ways. Trading off fewer use-after-free for longer build times and a more complex language may be a good tradeoff for the correctness of some projects, and a bad tradeoff for others.
> If GC is acceptable or you
BTW, a tracing GC - whose costs are now virtually entirely limited to a higher RAM footprint - is acceptable much more frequently than you may think. Sometimes, without being aware, languages like C, C++, Rust, or Zig may sacrifice CPU to reduce footprint, even when this tradeoff doesn't make sense. I would strongly recommend watching this talk (from the 2025 International Symposium on Memory Management), and the following Q&A about the CPU/footprint tradeoff in memory management: https://www.youtube.com/watch?v=mLNFVNXbw7I
Two things here, the easy one first: The RAM trade off is excellent for normal sizes but if you scale enormously the trade off eventually reverses. $100 per month for the outfit's cloud servers to have more RAM is a bargain compared to the eye-watering price of a Rust developer. But $1M per month when you own whole data centres in multiple countries makes hiring a Rust dev look attractive after all.
As I understand it this one is why Microsoft are rewriting Office backend stuff in Rust after writing it originally in C#. They need a safe language, so C++ was never an option, but at their scale higher resource costs add up quickly so a rewrite to a safe non-GC language, though not free, makes economic sense.
Second thing though: Unpredictability. GC means you can't be sure when reclamation happens. So that means both latency spikes (nasty in some applications) and weird bugs around delayed reclamation, things running out though you aren't using too many at once, etc. If your only resource is RAM you can just buy more, but otherwise you either have to compromise the language (e.g. a defer type mechanism, Python's "with" construct) or just suck it up. Java tried to fix this and gave up.
I agree that often you can pick GC. Personally I don't much like GC but it is effective and I can't disagree there. However you might have overestimated just how often, or missed more edge cases where it isn't a good choice.
I don't think you've watched the talk. The minimal RAM-per-core is quite high, and often sits there unused even though it could be used to reduce the usage of the more expensive CPU. You pay for RAM that you could use to reduce CPU utilisation and then don't use it. What you want to aim for is a RAM/CPU usage that matches the RAM/CPU ratio on the machine, as that's what you pay for. Doubling the CPU often doubles your cost, but doubling RAM costs much less than that (5-15%).
If two implementations of an algorithm use different amounts of memory (assuming they're reasonable implementations), then the one using less memory has to use more CPU (e.g. it could be compressing the memory or freeing and reusing it more frequently). Using more CPU to save on memory that you've already paid for is just wasteful.
Another way to think about it is consider the extreme case (although it works for any interim value) where a program, say a short-running one, uses 100% of the CPU. While that program runs, no other program can use the machine, anyway, so if you don't use up to 100% of the machine's RAM to reduce the program's duration, then you're wasting it.
As the talk says, it's hard to find less than 1GB per core, so if a program uses computational resources that correspond to a full core yet uses less than 1GB, it's wasteful in the sense that it's spending more of a more expensive resource to save on a less expensive one. The same applies if it uses 50% of a core and less than 500MB of RAM.
Of course, if you're looking at kernels or drivers or VMs or some sorts of agents - things that are effectively pure overhead (rather than direct business value) - then their economics could be different.
> Second thing though: Unpredictability. GC means you can't be sure when reclamation happens.
What you say may have been true with older generations of GCs (or even something like Go's GC, which is basically Java's old CMS, recently removed after two newer GC generations). OpenJDK's current GCs, like ZGC, do zero work in stop-the-world pauses. Their work is more evenly spread out and predictable, and even their latency is more predictable than what you'd get with something like Rust's reference-counting GC. C#'s GC isn't that stellar either, but most important server-side software uses Java, anyway.
The one area where manual memory management still beats the efficiency of a modern tracing GC (although maybe not for long) is when there's a very regular memory usage pattern through the use of arenas, which is another reason why I find Zig particularly interesting - it's most powerful where modern GCs are weakest.
By design or happy accident, Zig is very focused on where the problems are: the biggest security issue for low-level languages is out-of-bounds access, and Zig focuses on that; the biggest shortcoming of modern tracing GCs is arena-like memory usage, and Zig focuses on that. When it comes to the importance of UAF, compilation times, and language complexity, I think the jury is still out, and Rust and Zig obviously make very different tradeoffs here. Zig's bottom-line impact, like that of Rust, may still be too low for widespread adoption, but at least I find it more interesting.
> As I understand it this one is why Microsoft are rewriting Office backend stuff in Rust after writing it originally in C#
The rate at which MS is doing that is nowhere near where it would be if there were some significant economic value. You can compare that to the rate of adoption of other programming languages or even techniques like unit testing or code review. With any new product, you can expect some noise and experimentation, but the adoption of products that offer a big economic value is usually very, very fast, even in programming.
This totally ignores the role of memory bandwidth, which is often the key bottleneck on multicore workloads. It turns out that using more RAM costs you more CPU, too, because the CPU time is being wasted waiting for DRAM transfers. Manual memory management (augmented with optional reference counting and "borrowed" references - not the pervasive refcounting of Swift, which performs less well than modern tracing GC) still wins unless you're dealing with the messy kind of workload where your reference graphs are totally unpredictable and spaghetti-like. That's the kind of problem that GC was really meant for. It's no coincidence that tracing GC was originally developed in combination with LISP, the language of graph-intensive GOFAI.
Yes, memory bandwidth adds another layer of complication, but it doesn't matter so much once your live set is much larger than your L3 cache. I.e. a 200MB live set and a 100GB live set are likely to require the same bandwidth. Add to that the fact that tracing GCs' compaction can also help (with prefetching) and the situation isn't so clear.
> That's the kind of problem that GC was really meant for.
Given the huge strides in tracing GCs over the past ten and even five years, and their incredible performance today, I don't think it matters what those of 40+ years ago were meant for, but I agree there are still some workloads - not anything that isn't spaghetti-like, but specifically arenas - that are more efficient than tracing GCs (young-gen works a little like an arena but not quite), which is why GCs are now turning their attention to that kind of workload, too. The point remains that it's very useful to have a memory management approach that can turn the RAM you've already paid for to reduce CPU consumption.
Indeed, we're not seeing any kind of abandonment of tracing GC at a rate that is even close to suggesting some significant economic value in abandoning them (outside of very RAM-constrained hardware, at least).
That approach is specifically arenas: if you can put useful bounds on the maximum size of your "dead" data, it can pay to allocate everything in an arena and free it all in one go. This saves you the memory traffic of both manual management and tracing GC. But coming up with such bounds involves manual choices, of course.
It goes without saying that memory compaction involves a whole lot of extra traffic on the memory subsystem, so it's unlikely to help when memory bandwidth is the key bottleneck. Your claim that a 200MB working set is probably the same as a 100GB working set (or, for that matter, a 500MB or 1GB working set, which is more in the ballpark of real-world comparisons) when it comes to how it's impacted by the memory bottleneck is one that I have some trouble understanding also - especially since you've been arguing for using up more memory for the exact same workload.
Your broader claim wrt. memory makes a whole lot of sense in the context of how to tune an existing tracing GC when that's a forced choice anyway (which, AIUI, is also what the talk is about!) but it just doesn't seem all that relevant to the merits of tracing GC vs. manual memory management.
> we're not seeing any kind of abandonment of tracing GC at a rate that is even close to suggesting some significant economic value in abandoning them
We're certainly seeing a lot of "economic value" being put on modern concurrent GC's that can at least perform tolerably well even without a lot of memory headroom. That's how the Golang GC works, after all.
It doesn't go without saying that compaction involves a lot of memory traffic, because memory is utilised to reduce the frequency of GC cycles and only live objects are copied. The whole point of tracing collection is that extra RAM is used to reduce the total amount of memory management work. If we ignore the old generation (which the talk covers separately), the idea is that you allocate more and more in the young gen, and when it's exhausted you compact only the remaining live objects (which is a constant for the app); the more memory you assign to the young gen, the less frequently you need to do even that work. There is no work for dead objects.
> when it comes to how it's impacted by the memory bottleneck is one that I have some trouble understanding also - especially since you've been arguing for using up more memory for the exact same workload.
Memory bandwidth - at least as far as latency is concerned - is used when you have a cache miss. Once your live set is much bigger than your L3 cache, you get cache misses even when you want to read it. If you have good temporal locality (few cache misses), it doesn't matter how big your live set is, but the same is if you have bad temporal locality (many cache misses).
> which, AIUI, is also what the talk is about
The talk focuses on tracing GCs, but it applies equally to manual memory management (as discussed in the Q&A; using less memory for the same algorithm requires CPU work regardless if it's manual or automatic)
> when that's a forced choice
I don't think tracing GCs are ever a forced choice. They keep getting chosen over and over for heavy workloads on machines with >= 1GB/core because they offer a more attractive tradeoff than other approaches for some of the most popular application domains. There's little reason for that to change unless the economics of DRAM/CPU change significantly.
It definitely tracks with my experience. Did you see Chrome on AMD EPYC with 2TB of memory? It reached like 10% of Mem utility but over 46% of CPU around 6000 tabs. Mem usage climbed steeply at first but got overtaken by CPU usage.
All you know is that CPU is exhausted before the RAM is, which, if anything, means that it may have been useful for Chrome to use more RAM (and reduce the liveset-to-heap ratio) to reduce CPU utilisation, assuming this CPU consumption has anything to do with memory management.
There is no situation in which, given the same allocation rate and live set, adding more heap to a tracing GC makes it work more. That's why in the talk he says that a DIMM is a hardware accelerator for memory management if you have a tracing-moving collector: increase the heap and voila, less CPU is spent on memory management.
That's why tracing-moving garbage collection is a great choice for any program that spends a significant amount of CPU on memory management, because then you can reduce that work by adding more RAM, which is cheaper than adding more CPU (assuming you're running on a machine that isn't RAM-constrained, like small embedded devices).
I've spent much of my career working on desktop software, especially on Windows, and especially programs that run continuously in the background. I've become convinced that it's my responsibility to treat my user's machines as RAM-constrained, and, outside of any long-running compute-heavy loops, to value RAM over CPU as long as the program has no noticeable lag. My latest desktop product was Electron-based, and I think it's pretty light as Electron apps go, but I wish I'd had the luxury of writing it all in Rust so it could be as light as possible (at least one background service is in Rust). My next planned desktop project will be in Rust.
A recent anecdote has reinforced my conviction on this. One of my employees has a PC with 16 GB of RAM, and he couldn't run a VMware VM with 4 GB of guest RAM on that machine. My old laptop, also with 16 GB of RAM, had plenty of room for two such VMs. I didn't dig into this with him, but I'm guessing that his machine is infested with crap software, much of which is probably using Electron these days, each program assuming it can use as much RAM as it wants. I want to fight that trend.
The part you've mistaken for being somehow general and relevant to our conversation is about trade-offs within GC. So it's completely irrelevant rather than being, as you seemed to imagine, an important insight. It actually reminds me of the early 1940s British feedback on intelligence for the V2 rocket. British and American Scientists were quite sure that Germany could not develop such a rocket, toy rockets work but at scale this rocket cannot work.
You can buy those toys today, a model store or similar will sell you the basics so you can see for yourself, and indeed if you scale that up it won't make an effective weapon. However intelligence sources eventually revealed how the German V2 was actually fuelled. To us today having seen space rockets it's obvious, the fuel is a liquid not a solid like the toy, which makes fuel loading rather difficult but delivers enormously more thrust. The experts furiously recalculated and discovered that of course these rockets would work unlike the scaled up toy they'd been assessing before. The weapon was very real.
Anyway. The speaker is assuming that we're discussing how much RAM to use on garbage. Because they're assuming a garbage collector, because this is a talk about GC. But a language like Rust isn't using any RAM for this. "The fuel is liquid" isn't one of the options they're looking at, it's not what their talk is even about so of course they don't cover it.
> With any new product, you can expect some noise and experimentation, but the adoption of products that offer a big economic value is usually very, very fast, even in programming.
What you've got here is the perfect market fallacy. This is very, very silly. If you're young enough to actually believe it due to lack of first hand experience then you're going to have a rude awakening, but I think sadly it probably just means you don't like the reality here and are trying to explain it away.
The speaker is one of the world's leading experts on memory management, and the "mistake" is one of the biggest breakthroughs in software history, which is, today, the leading chosen memory management approach. Tony Hoare has done this when his mistakes became apparent; it's hard to find people who say "it was a terrible idea and I regret it" when the idea has won spectacularly and is widely recognised to be quite good.
> Anyway. The speaker is assuming that we're discussing how much RAM to use on garbage. Because they're assuming a garbage collector, because this is a talk about GC. But a language like Rust isn't using any RAM for this. "The fuel is liquid" isn't one of the options they're looking at, it's not what their talk is even about so of course they don't cover it.
Hmm, so you didn't really understand the talk, then. You can reduce the amount of garbage to zero, and the point still holds. To consume less RAM - by whatever means - you have spend CPU. After programming in C++ for about 25 years, this is obvious to me, as it should be to anyone who does low-level programming.
The point is that a tracing-moving algorithm can use otherwise unused RAM to reduce the amount of CPU spent on memory management. And yes, you're right, usually in languages like C++, Zig, or Rust we typically do not use extra RAM to reduce the amount of CPU we spend on memory management, but that doesn't mean that we couldn't or shouldn't.
> What you've got here is the perfect market fallacy.
A market failure/fallacy isn't something you can say to justify any opinion you have that isn't supported by empirical economics. A claim that the market is irrational may well be true, but it is recognised, even by those who make it, as something that requires a lot of evidence. Saying that you know what the most important thing is and that you know the best way to get it, and then use the fact that most experts and practitioners don't support your position as evidence that it is correct. That's the Galileo complex: what proves I'm right is that people say I'm wrong. Anyway, a market failure isn't something that's merely stated; it's something that's demonstrated.
BTW, one of those times Tony Hoare said he wrong? It was when he claimed the industry doesn't value correctness or won't be able to achieve it without formal proofs. One of the lesssons from that in the software correctness community is to stop claiming or believing we have found the universally best path to correctness, and that's why we stopped doing that in the nineties. Today it's well accepted there can be many effective paths to correctness, and the research is more varied as well.
I started programming in 1988-9, I think, and there's been a clear improvement in quality even since then (despite the growth in size and complexity of software). Rust makes me nostalgic because it looks and feels and behaves like a 1980s programming language - ML meets C++ - and I get its retro charm (and Zig has it, too, in its Sceme meets C sort of way), but we've learnt, what Tony Hoare has learnt, is that there are many valid approaces to correctness. Rust offers one approach, which may be attractive to some, but there are others that may be attractive to more.
Nah, in the model you're imagining now the program takes infinite time. But we can observe that our garbage free Rust program doesn't take infinite time, it's actually very fast. That's because your model is of a GC system - where ensuring no garbage really would need infinite time and a language without GC isn't a GC with zero garbage, it's entirely different, that's the whole point.
More generally, a GC-less solution may be more CPU intensive, or it may not, and although there are rules of thumb it's difficult to have any general conclusions. If you work on Java this is irrelevant, your language requires a GC, so this isn't even a question and thus isn't worth mentioning in a talk about, again, the Java GC.
> A claim that the market is irrational may well be true
Which makes your entire thrust stupid. You depend upon the perfect market fallacy for your point there, the claim that if this was a good idea people would necessarily already be doing it - once you accept that's a fallacy you have nothing.
Wat? Where did you get that?
> That's because your model is of a GC system - where ensuring no garbage really would need infinite time and a language without GC isn't a GC with zero garbage, it's entirely different, that's the whole point.
Except it's not different, and working "without garbag"e doesn't take infinite time even with a garbage collector. For example, a reference counting GC also has zero garbage (in fact, that's the GC Rust uses to manage Rc/Arc) and doesn't take infinite time. It does, however, sacrifice CPU for lower RAM footprint. Have you studied the theory of memory management? The CPU/memory tradeoff exists regardless of what algorithm you use for memory management, it's just that some algorithms allow you to control that tradeoff within the algorithms and others require you to commit to a tradeoff upfront.
For example, using an arena in C++/Rust/Zig is exactly about exploiting that tradeoff (Zig in particular is very big on arenas) to reduce the CPU spent on memory management by using more RAM: the arena isn't cleared until the transaction is done, which isn't minimal in terms of RAM, but requires less CPU. Oh, and if you use an arena (or even a pool), you have garbage, BTW. Garbage means an object that isn't used by the program, but whose memory cannot yet be reused for the next allocation.
If you think low-level languages don't have garbage, then you haven't done enough low-level programming (and learn about arenas; they're great). There are many pros and many cons to the Rust approach, and it sure is a good tradeoff in some situations, but the reason the biggest Rust zealots - those who believe it's universally superior - are those who haven't done much low-level programming and don't understand the tradeoffs it involves. It's also them who think that the reason those of us who were there first picked up C++ and later abandoned it for some use-cases did so only because it wasn't memory-safe. That was one reason, but there were many others, at least as equally decisive. Rust fixes some of C++'s shortcomings but certainly not all; Zig fixes others (those that happen to be more important to me) but certainly not all. They're both useful, each in their own way, and neither comes close to being "the right way to program". Knowledgeable, experienced people don't even claim that, and are careful to point out that they may genuinely believe some universal superiority but that they don't actually have the proof.
Whether you use a GC (tracing-moving as in Java, refcounting as in Rust's Rc or Swift, or tracing-sweeping as in Go), use arenas, or manually do malloc/free, the same principles and tradeoffs of memory management apply. That's because abstract models of computation - Turing machines, the lambda calculus, or the game of life, pick your favourte - have infinite memory, but real machines don't, which means we have to reuse memory, and doing that requires computational work. That's what memory management means. Some algorithms, like Rust's primitive refcounting GC, aim to reuse memory very aggressively, which means they do some work (such as updating free-lists) as soon as some object is unused so that its memory can be reused immediately, while other approaches postpone the reuse of memory to do less work. That's what tracing collectors or arenas do, and that's precisely why we who do low-level programming like arenas so much.
Anyway, the point of the talk is this: the CPU/memory tradeoff is, of course, inherent to all approaches of memory management (though not necessarily in every particular use case), and so it's worth thinking about the fact that the minimal amount of RAM per core on most machines these days is high. This applies to everything. Then he explains that the trancing-moving collectors allow you to turn the tradeoff dial - within limits - within the same algorithm. Does it mean tracing-moving collection is always the best choice? No. But do approaches that strive to minimise footprint somehow evade the CPU/memory tradeoff? Absolutely not.
A lesson that a low-level programmer may take away from that could be something like, maybe I should rely on RC less and, instead, try to use larger arenas if I can.
> Which makes your entire thrust stupid. You depend upon the perfect market fallacy for your point there
Except I don't depend on it. I say it's evidence that cannot be ignored. The market can be wrong, but you cannot assume it is.
The talk you're so excited about actually shows this asymptote. In reality a GC doesn't actually want zero garbage because we're trading away RAM to get better performance. So they don't go there, but it ought to have pulled you up short when you thought you could apply this understanding to an entirely unrelated paradigm.
Hence the V2 comparison. So long as you're thinking about those solid fuel toy rockets the V2 makes no sense, such a thing can't possibly work. But of course the V2 wasn't a solid fuel rocket at all and it works just fine. Rust isn't a garbage collected language and it works just fine.
GC is very good for not caring about who owns anything. This can lead to amusing design goofs (e.g. the for-each bug in C# until C# 5) but it also frees programmers who might otherwise spend all their time worrying about ownership to do something useful which is great. However it really isn't the panacea you seem to have imagined though at least it is more widely applicable than arenas.
> it's worth thinking about the fact that the minimal amount of RAM per core on most machines these days is high.
Like I said, this means you are less likely to value the non-GC approach when you're small. You can put twice as much RAM in the server for $50 so you do that, you do not hire a Rust programmer to make the software fit.
But at scale the other side matters - not the minimum but the maximum, and so whereas doubling from 16GB RAM to 32GB RAM was very cheap, doubling from 16 servers to 32 servers because they're full means paying twice as much, both as capital expenditure and in many cases operationally.
> I say it's evidence
I didn't see any evidence. Where is the evidence? All I saw was the usual perfect market stuff where you claim that if it worked then they'd have already completed it by some unspecified prior time, in contrast to the fact I mentioned that they've hired people to do it. I think facts are evidence and the perfect market fallacy is just a fallacy.
Oh, I see the confusion. That asymptote is for the hypothetical case where the allocation rate grows to infinity (i.e. remains constant per core and we add more cores) while the heap remains constant. Yes, with an allocation rate growing to infinity, the cost of memory management (using any algorithm) also grows to infinity. That it's so obvious was his point showing why certain benchmarks don't make sense as they increase the allocation rate but keep the heap constant.
> So they don't go there, but it ought to have pulled you up short when you thought you could apply this understanding to an entirely unrelated paradigm.
I'm sorry, but I don't think you understand the theory of memory management. You obviously run into these problems even in C. If you haven't then you haven't been doing that kind of programming long enough. Some results are just true for any kind of memory management, and have nothing to do with a particular algorithm. It's like how in computational complexity theory, certain problems have a minimal cost regardless of the algorithm chosen.
> But at scale the other side matters - not the minimum but the maximum, and so whereas doubling from 16GB RAM to 32GB RAM was very cheap, doubling from 16 servers to 32 servers because they're full means paying twice as much, both as capital expenditure and in many cases operationally.
I've been working on servers in C++ for over 20 years, and I know the tradeoffs, and when doing this kind of programming seeing CPU exhausted before RAM is very common. I'm not saying there are never any other situations, but if you don't know how common this is, then it seems like you don't have much experience with low-level programming. Implying that the more common reason to need more servers is because what's exhausted first is the RAM is just not something you hear from people with experience in this industry. You think that the primary reason for horizontal scaling is dearth of RAM?? Seriously?! I remember that in the '90s or even early '00s we had some problems of not enough RAM on client workstations, but it's been a while.
In the talk he tells memory-management researchers about the economics in industry, as they reasonably might not be familiar with them, but decision-makers in industry - as a whole - are.
Now, don't get me wrong, low level languages do give you more control over resource tradeoffs. When we use those languages, sometimes we choose to sacrifice CPU for footprint and use a refcounting GC or a pool when it's appropriate, and sometimes we sacrifice footprint for less CPU usage and use an arena when it's appropriate. This control is the benefit of low level languages, but it also comes at a significant cost, which is why we use such languages primarily for software that doesn't deliver direct business value but is "pure overhead", like kernels, drivers, VMs, and browsers, or for software running on very constrained hardware.
> I didn't see any evidence.
The evidence is that in a highly competitive environment of great economic significance, where getting big payoffs is a way to get an edge over the competition, technologies that deliver high economic payoffs spread quickly. If they don't, it could be a case of some market failure, but then you'd have to explain why companies that can increase their profits significantly and/or lower their prices choose not to do so.
When you claim some technique would give a corporation a significant competitive edge, and yet most corporations don't take it (at least not for most projects), then that is evidence against that claim because usually companies are highly motivated to gain an advantage. I'm not saying it's a closed case, but it is evidence.
Corporations will generally want to optimize for ease of development and general ecosystem maturity. Rust is at a clear disadvantage at least wrt. the latter - the first generally usable version of the language was only released in late 2018 - and other safe languages are generally GC-based (compare Java/C# with C++). It's quite normal that a lot of performance would be left on the table wrt. both CPU and memory footprint.
A big factor is that Java was the entire ecosystem, you're getting a programming language (which is pretty good), a novel virtual machine (most of their ideas fell through but the basic thing is fine), dedicated hardware (mostly now forgotten), a component architecture, and a set of tools for them.
You're still on this 1GB/core thing which is the wrong end of the scale in two senses. Firstly, I've worked on systems where we'd want 1TB/core and so today that means you're buying a lot of CPU performance you don't need to get enough RAM because as you say, that machine is "exhausted" anyway.
But more importantly the scale for big products isn't dictated by RAM or CPU it's dictated by service provision, and at large scale that's just linear. Twice as much provision, twice the cost. Avoiding a GC can let you slash that cost. Cutting a $1M annual bill in half would justify hiring a Rust programmer though likely not a whole team. Cutting a $1Bn annual bill in half - which is much more like what Microsoft are spending on O365 - is obviously worth hiring a team.
It's not instant. GC tuning is basically instant. RIIR might take three, five, even ten years. So don't expect results tomorrow afternoon.
> Avoiding a GC can let you slash that cost
But it doesn't, because RAM isn't the bottleneck in the vast majority of cases. It doesn't matter how linear costs are if RAM isn't the thing that's exhausted. That's why the use of manual memory management had been declining for decades.
At 1TB per core things still don't change because GC no longer has a high footprint cost in the old gen. You may use 15x RAM for the first 50 MB, but the overhead for the second GB is very small, thanks to the generational hypothesis: the older an object is, the less frequently it is allocated. The cost of a moving-tracing GC is propprtional to allocation-rate * live-set / heap-size per generation. When the live set is large, the allocation rate in that generation is low, which means that the heap size premium is also low.
> So don't expect results tomorrow afternoon.
Manual memory management is the older option, and it's been in decline for decades precisely because the savings in costs go in the other direction (not in all situations, but in most) due to the economics of RAM costs vs CPU costs. Without a marked change in the economics, the trend won't reverse even in another 50 years.
...to the extent possible within their project budget. Otherwise the product would — as GP already pointed out — not exist at all, because the project wouldn't be undertaken in the first place.
> This goal can be reached by learning to write Rust; practice makes perfect.
Pretty sure it could (at least) equally well be reached by learning to write Ada.
This one-note Rust cult is really getting rather tiresome.
Sure, but when other engineers discover that shit caused us many defects (e.g., asbestos as a fire insulator), they don't turn around and say, "Well, asbestos sure did cause a lot of cancer, but Cellulose Fibre doesn't shield us from neutron radiation. So it won't be preventing all cancers. Ergo, we are going back to asbestos."
And then you have team Asbestos and team Lead paint quarrelling who has more uses.
That's my biggest problem. The cyclic, Fad Driven Development that permeates software engineering.
> Pretty sure it could (at least) equally well be reached by learning to write Ada.
Not really. Ada isn't that memory safe. It mostly relies on runtime checking [1]. You need to use formal proofs with Ada SPARK to actually get memory safety on par with Rust.
> Pretty sure it could (at least) equally well be reached by learning to write Ada.
See above. You need Ada with Spark. At that point you get two files for each method like .c/.h, one with method definition and one with proof. For example:
// increment.ads - the proof
procedure Increment
(X : in out Integer)
with
Global => null,
Depends => (X => X),
Pre => X < Integer'Last,
Post => X = X'Old + 1;
// increment.adb - the program
procedure Increment
(X : in out Integer)
is
begin
X := X + 1;
end Increment;
[1] https://ajxs.me/blog/How_Does_Adas_Memory_Safety_Compare_Aga...
The problem with this statement is that without a memory safety invariant your code doesn't compose. Some code might assume no UAF and other parts could and you'd have a mismatch. Just like borrow checker is viral, so is the unsafety.
> If you don't say what Zig programmers should give up to gain more safety, saying "all new languages should be memory-safe" is about as meaningful as saying we should write fewer bugs.
The goal of all engineering disciplines, including software, should be a minimization of errors and defects.
Here is how engineering in any other non-computer science field takes place. You build something. See where it breaks; try to build it again given time and budget constraints. Eventually you discover certain laws and rules. You learn the rules and commit them to a shared repository of knowledge. You work hard to codify those laws and rules into your tools and practice (via actual government laws). Furthermore, you try to build something again, with all the previous rules, tools, and accumulated knowledge.
How it works in tech. You build something. See where it breaks, say that whoever built it was a cream-for-brain moron and you can do it better and cheaper. Completely forget what you learned building the previous iteration. See where it breaks. Blame the tools for failure; remove any forms of safety. Project cancelled due to excessive deaths. Bemoan the lack of mental power in newer hires or lack of mental swiftness in older hires. Go to step 1.
You'll notice a stark contrast between Engineering and Computer Tech. Computer tech is pop culture. It's a place where people wage wars about whether lang X or lang Y is better. How many times did programming trend go from static to dynamically typed? How many times did programming learned a valuable lesson, only for everyone to forget it, until decades later another language resurrected it?
Ideally, each successive language would bring us closer and closer to minimizing defects, with more (types of) safety and better guarantees. Is Rust a huge leap compared to Idris? No, it's better than Ada at memory safety that's for sure.
But it's managed to capture a lot of attention, and it is a much stricter language than many others. It's a step towards ideal. And how do programmers react to it? With disgust and a desire for less safety.
Sigh. I guess we deserve all the ridicule we can get.
Yes, but that holds for any correctness property, not just the 0.0001% of them that memory safe languages guarantee. That's why we have bugs. The reason memory safety is a focus is because out-of-bounds access is the leading cause of dangerous vulnerabilities.
> The goal of all engineering disciplines, including software, should be a minimization of errors and defects.
Yes, but practical minimisation, not hypothetical minimisation, i.e. how can I get the least bugs while keeping all my constraints, including budget. Like I said, a language like Rust exists because minimisation of errors is not the only constraints, because if it were, there are already far more popular languages that do just as much.
> You'll notice a stark contrast between Engineering and Computer Tech.
I'm not sure I buy this, because physical, engineered objects break just as much as software does, certainly when weighted by the impact of the failure. As to learning our lessons, I think we do when they are actually real. Software is a large and competitive economic activity, and where's there's a real secret to more valuable software, it spreads like wildfire. For example, high-level programming languages spread like wildfire; unit tests and code review did, too. And when it comes to static and dynamic typing, the studies on the matter were inconclusive except in certain cases such as JS vs TS; and guess what? TS has spread very quickly.
The selective pressures are high enough, and we see how well they work frequently enough that we can actually say that if some idea doesn't spread quickly, then it's likely that its impact isn't as high as its fans may claim.
> And how do programmers react to it? With disgust and a desire for less safety.
I don't think so. In such a large and competitive economic activity, the assumption that the most likely explanation to something is that the majority of practitioners are irrational seems strange to me. Rust has had some measure of adoption and the likeliest explanation for why it doesn't have more is the usual one for any product: it costs too much and delivers too little.
Let's say that the value, within memory safety, between spatial and temporal safety is split 70-30; you know what? let's say 60-40. If I can get 60% of Rust's value for 10% of Rust's cost, that a very rational thing to do. I may even be able to translate my savings into an investment in correctness that is more valuable than use-after-free.
Rust achieves practical minimization, if not outright eradication, of a set of errors even in practice. And not just memory safety errors.
> Like I said, a language like Rust exists because minimisation of errors is not the only constraints, because if it were, there are already far more popular languages that do just as much.
The reason Rust exists is that the field hasn't matured enough to accept better engineering practices. If everyone could write and think in pre/post/invariant way, we'd see a lot fewer issues.
> I'm not sure I buy this, because physical, engineered objects break just as much as software does, certainly when weighted by the impact of the failure.
Dude, the front page was about how Comet AI browser can be hacked by your page and ordered to empty your bank account. That's like your fork deciding to gut you like a fish.
> the assumption that the most likely explanation to something is that the majority of practitioners are irrational seems strange to me.
Why? Just because you are intelligent doesn't mean you are rational. Plenty of smart people go bonkers. And looking at the state of the field as a whole, I'd have to ask for proof it's rational.
It achieves something in one way, while requiring you to pay some price, while other languages achieve something in a different way, with a different cost. I've been involved with software correctness for many, many years (early in my career I worked on safety-critical, hard realtime software, and then practised and taught formal methods for use in industry), and there is simply no research, none, suggesting that Rust's approach is necessarily the best. Remember that most software these days - not the kind that flies aeroplanes or controls pacemakers, that's written mostly in C, but the kind that runs your bank, telecom supplier, power company, healthcare etc. but also Facebook - is written in languages that offer the same guarantees as Rust, for better or worse.
> If everyone could write and think in pre/post/invariant way, we'd see a lot fewer issues.
Except I've worked with formal methods in a far more rigorous way than Rust offers (well, it offers almost nothing), and in this field there is now the acknowledgement that software correctness can be achieved in many different ways. In the seventies there was a consensus about how to write correct software, and then in the nineties it all got turned around.
> That's like your fork deciding to gut you like a fish.
I don't think so, because everyone uses a fork but this is the first time I hear about Comet AI. The most common way to empty people's bank accounts is still by conning them.
> Why? Just because you are intelligent doesn't mean you are rational.
Intelligence has nothing to do with it. But "rational" here means "in accordance with reality", and since software is a major competitive economic activity, which means that if some organisations act irrationally, there's both a strong motivation and an ability to take their lunch money. If they still have it, they're probably not behaving irrationally.
Pet projects are nice, but slow down with the copium intake.
The future is many things, but a love letter to C is definitely not it.
Zig is cute and a fun hobby project that might see above average success for a hobby project. But that's about it. It doesn't address the problems people actually have with C++, not like Rust or Swift do, and it certainly isn't going to attract any attention from the Java, JavaScript, C#, Python, etc... of the world.
> It doesn't address the problems people actually have with C++, not like Rust or Swift do
Speak for yourself. Zig addresses pretty much all of my problems with C++:
- terrible compile times - overly complex and still underpowered template/meta programming - fragmented/ancient build tools - actually good enums/tagged unions - exceptions - outdated OOP baggage
I don’t actually mind C++ that much, but Zig checks pretty much all of my boxes. Rust/swift check some of these but not all and add a few of their own.
> and it certainly isn't going to attract any attention from the Java, JavaScript, C#, Python, etc... of the world.
Yeah of course. Zig isn’t trying to be a max ergonomics scripting language…
That doesn't seem that odd to me. It's a trade off: more flexibility, but more manual work. Maybe I have a buffer that I've allocated that I'm not using anymore (say I have a buffer pool) and want to use it again. If the type allocates its own behind the scenes, I can't do that. Or maybe I'm working in an environment where I need to statically allocate all of my resources up-front, and can't allocate later.
The big downside is that if 90% of people are just going to allocate a buffer and pass it in, it sucks that 90% of people need to do more work and understand more minutiae when only 10% of the people actually need to. The holy grail is to give lots of flexibility, but make the simple/common case easy.
A simple improvement to this interface might be to allow the caller to pass a zero-length buffer (or Zig's version of null), and then the type will allocate its own buffer. Of course, there's still a documentation burden so people know they can do that. Another option could be to have second constructor function that takes no buffer arguments at all, which allocates the buffers and passes them to the fully-flexible constructor function.
Isn't that the reason why Zig passes around allocators everywhere? If you're using a buffer pool, you should probably be handing out some kind of buffer pool allocator.
Requiring all allocation to have happened before execution is still a good reason to pass buffers around, but I feel like the other situations you describe can be solved by just passing the right allocators.
For those working on the standard library, it’s a great thing. For one like me who casually uses zig, it feels like waiting for 0.16.0 for most of the IO dust to settle is the right thing to do.
The future looks bright after that though. Both Andrew and Loris seem to think this is the last major breaking change, so I hope we can see a 1.0 not too long after. The biggest uncertainty now is what the consequences of (re-)introducing stack-less coroutines will be.
Well I mean the standard library still needs a _lot_ of work IMHO. Process management for instance is at a stage where it's barely usable for simple use cases.
This pattern (new language evolves to be as complex as the languages it was supposed to replace) seems familiar.
Such approach is not a great fit for those who treat Zig as a future job opportunity, but for personal and small-team projects it's _already_ a neat language with clear goals and great tooling.
Since 0.15, though, I feel too dumb for Zig’s ArrayList.
My info could be outdated - I don't follow Zig very closely, but I am curious.
No promises about potential future changes though :^)
That made me think of how that change would be received in Go (probably would be discarded). They way they approach changes in extremely deep analysis and taking as much time as it needs to avoid mistakes and reach a consistent solution (or as close as possible).
This has been my favorite for a while: https://github.com/golang/go/issues/45624
4 years to decide on something relatively minor, that right now can be done with a bit of a one-liner extra work. But things need to be well thought out. Inconsistencies are pointed out. Design concerns are raised. Actual code usage in the real world are taken into account... too slow for some people, but I think it's just as slow as it needs to be. The final decision is shaping out to be very nice.
But when rust ships features to stable, they’re usually pretty well thought through. I’m impatient. But the rust language & compiler teams probably have the right idea.
If we all stopped complaining about memory safety in C and C++, we would never have gotten rust in the first place. Rob pike would never have let generics make it into Go if people didn’t spend years pestering him about it.
I think I fail by taking for granted all the work that’s gone into languages like rust. Lots of smart people have poured themselves into these discussions and designs. Thankyou. I don’t say that enough.
But yeah, async rust seems like one of those places where people arguing it out on github didn’t converge at a great design. It’s a really hard problem - really, combining a borrow checker with stack suspension is a CS research problem. I think it’s interesting and important to acknowledge that committees don’t always do great research. Sometimes you need a bunch of iterations of an idea. And a bright cookie or two who can work without justifying their designs.
So no, I don’t think I’ll stop talking about it. Making async rust v1 was an incredible amount of work and I’m very grateful. But I also really want to seed the idea space so some bright sparks can think about what async rust v2 might look like.
It works now that way because they promised it to be stable after 1.0.
I think fact that Zig can be used as a C/C++ cross compiler is brilliant.
For a lot of simple calls, that works out pretty well, once you know all the tricks to Zig's syntax. A lot of requirements and implications are generally written out in very simple and short functions that are often logically named. Things like Allocators are pretty easy conceptually even if you probably don't want to write one yourself.
It all breaks down when you start dealing with complex concepts. Zig's new I/O system looks a lot like Java's streams and wrappers and readers and writers, all wrapping around each other to make sending encrypted text over a secure channel as simple as output.write("hello").
I think the new I/O system combined with the lack of documentation about how to use it was a mistake. I'm not even sure if expressing a typing system as complicated as this in the Zig standard library is a good idea. The entire language runs on clear, concise, short and readable methods, and the new system doesn't seem idiomatic in that way.
Am I missing some context? I’d love to hear it.
Discriminated unions, error handling, comptime, defer.
Better default integer type casting, ability to choose between releaseSafe/releaseFast
And probably other things.
As for comparison to Rust, you do want very low level memory handling for writing databases as an example. It is extremely difficult to write low level libraries in Rust
Did you write any Zig code yet? In terms of enforced correctness in the language (e.g. no integer promotion, no implicit 'dangerous' casts, null-safety, enforced error handling, etc...) and runtime safety (range-, nullptr-, integer-overflow-checks etc...), Zig is much closer to Rust than it is to C and C++.
It "just" doesn't solve static memory safety and some (admittedly important) temporal memory safety issues (aka "use-after-free"), but it still makes it much harder to accidentially trigger memory corruption as a side effect in most situations that C and C++ let slip through via a mix of compile errors and runtime checks (and you get ASAN/UBSAN automatically enabled in debug builds, a debug allocator which detects memory leaks and use-after-free for heap-allocations (unfortunately not for stack allocations), and proper runtime stack traces - things that many C/C++ toolchains are still missing or don't enable by default).
There is still one notable issue: returning a reference to stack memory from a function - this is something that many unexperienced Zig programmers seem to stumble into, especially since Zig's slice syntax looks so 'innocent' (slices look too similar to arrays, but arrays are values, while slices are references - e.g. 'fat pointers') - and which IMHO needs some sort of solution (either a compile time error via watertight escape analysis, or at least some sort runtime check which panics when trying to access 'stale' data on the stack) - and maybe giving slices their own distinct syntax that doesn't overlap with arrays might also help a bit.
But there is still the "big stuff" - the things that have a fundamental, architectural impact. Things like: Will my program be multithreaded? Will I have many systems that interact? Will my program be maximally memory-efficient? Do I have the capacity (or money) to ensure correctness if I say "yes" to any of that?
The most important consideration in any software project is managing architectural complexity. Zig is better, yes, but not a paradigm shift. If you say "yes" to any of the above, you are in the same world of pain (or expenses) as you would be in C or C++. This is the reason that Rust is interesting: It makes things feasible/cheap that were previously very hard/expensive, at a fundamental level.
But it doesn't have to be and it shouldn't. Rust also isn't a paradigm shift, it "just" solved static memory safety (admittedly a great engineering feat) but other languages solved memory safety too decades ago, just with more of it happening at runtime.
But in many other areas Rust copied too many bad ideas from C++ (and many of those "other things" Zig already does much better than Rust - but different people will have vastly different opinions about whether one solution is actually better than another - so discussions about those features usually run in circles).
There shouldn't be a single language that solves all problems - this will just result in stagnation, it's much better to have many languages to pick from - and even if they just differ in personal opinions of the language author of how to do things or entirely subjective 'syntax sugar' features. Competition is good, monocultures are bad.
> The most important consideration in any software project is managing architectural complexity
No language will help you managing "architectural complexity" in any meaningful way except by imposing tons of arbitrary restrictions which then get in the way for smaller projects that don't need much "architecture". We have plenty of "Enterprise-scale languages" already (Java, C#, Rust, C++, ...), what we need is more languages for small teams that don't get in the way of just getting shit done, since the really interesting and innovative stuff doesn't happen in enterprise environments but in small 'basement and bedroom teams' :)
Keep in mind that Rust's definition of "memory safety" covers much more than just use-after-free, the most important being thread safety. It is a blanket guarantee of no undefined behavior in any code that doesn't contain the word `unsafe`. Undefined behavior, including data race conditions, is a major time sink in all non-hobby C or C++ projects.
What bad ideas from C++ did Rust copy, in your opinion? I'm really not sure what you mean. Smart pointers? RAII?
There are plenty of languages that enable quick iteration, prototyping, or "just getting shit done". If that's what you need, why not use them? I'm personally more concerned about the finished product.
...I don't use them as "scare quotes", it's more like in "can't you just..." - e.g. something that looks simple from the outside but is hard to do / complex on the inside - e.g. I do recognize the work that went into Rust's memory safety, but I question the effort that went into it compared to more traditional memory safety methods, especially when looking at the restrictions that Rust imposed on the programmer - it's a pretty hefty tradeoff (IMHO of course).
> What bad ideas from C++ did Rust copy, in your opinion?
Mainly doing things in the stdlib that should be built into the language (e.g. Option, Result, Box, Cell, RefCell, Arc, and probably a dozen other types and traits...), resulting in what I call 'bird droppings syntax' of too many chained function calls to get to the thing you actually want (.unwrap, .into_iter, .iter, .iter_mut, .as_ref, .as_mut, .expect, .unwrap_or_else, .unwrap_or_default, .ok, .ok_or_else, .and_then, .or_else ... like, wtf?). The absurd amount of `::` and `<>` in typical Rust code. The missing separation line between stdlib and language (like for-loops using the Iterator trait, or more obviously: operator overloading). The stdlib doing memory allocations behind your back through a global allocator - which we know from C++ do be a really bad idea for decades already... etc etc... I think most Rust programmers are blind towards those issues the same way that C++ programmers are blind towards C++ issues (which is another thing both language ecosystems have in common though).
I don't thing there is any substantial difference between `Option<Thing>` and `@Nullable Thing` or `Thing | null`, I don't think there's anything wrong with choosing `::` over `.` for namespace resolution (it means you can have local variables with the same name as a module), and you have to have some way to declare generic parameters.
Rust generally does not allocate behind your back, but custom allocators is a work in progress. The reason it takes time is precisely that they want to avoid the mistakes of C++. A couple of mistakes were already avoided - for example, async/await in Rust does not allocate behind your back, while C++ coroutines do.
I object to @Nullable and similar arrangements as a magic special case. If the sum types only solved this one issue then it's a wash but they do a lot more too.
Either of the sum types, the concrete `Option<Thing>` or the ad hoc `Thing | null` are OK because they're not magic, less magic is better.
Technically Option is actually magic, though it's for a subtle reason unconnected to the current topic. If you go read its source Option is a langitem, that is, Rust literally isn't allowed to exist without this type. Rust's core libraries are defined, so you shouldn't and most people never will use Rust without them, but the Rust language actually doesn't require that all of core exists, it does require a handful of types, functions etc. and those are called "langitems".
So, why is Option a langitem? Well, Rust's language has for-each loops. But what it actually does (there's literally a compiler step doing this) is treat those loops as if they'd been written as a slightly clunky loop { } block making and then consuming an Iterator. To do that the IntoIterator and Iterator traits must exist, and further as the return type from the next call needed in Iterator is an Option the Option generic type must exist too!
Technically this type wouldn't have to be called Option, the Rust compiler doesn't care what it is called, but it must exist or else the compiler doesn't know how a for-each loop can work.
Another example is the `Deref` trait, which roughly corresponds to `operator->()` or `operator*()`, or an implicit reference conversion. It is called implicitly so you can use `smart_ptr.foo` without special syntax for dereferencing.
I think I fundamentally don't understand why any of this is a problem?
I don't like @Nullable (or the ? syntax used in C# for example) because they're a special case magic. They handle exactly one narrow idea (Tony's Billion Dollar Mistake, the "null" pointer) and nothing else.
I prefer sum types - both Option<T> and T | null are syntax for sum types. The former, which Rust has, is an explicit generic type, while the latter is an ad hoc typing feature. I don't have a strong preference between them.
What are you saying? Do you think I was sent here by some sinister cabal that organizes an effort to direct any discussion about programming towards Rust?
This has been not my experience at all in the ~6 years I've been writing Zig. I started having very little experience writing C (<1000, lines all written while in university) and since day 1 Zig has been a tremendous improvement over it, even back when it was at version 0.4.0.
I'm informed by having shipped a lot of C++ code in my time, which has taught me a lot about actually delivering stable software. Being better than C is a very low bar here.
There is plenty of working software written with pretty much any language
Really? I've not found it at all difficult to write low level libraries in Rust.
Also using arena and other special allocators in different sections of the program. While maintaining hard memory limits for different sections of the program.
These are possible to do in rust but it is very difficult for me so I decided to just not do it. Otherwise I have to spend 5x the normal amount of time to make sure the library cannot be misused by the user ever.
This is pretty pointless since I’m the only one who is going to use that code anyway.
Could be skill issue or w/e but I just find zig easier to make progress in
People do systems programming in rust, but that's not really what most of the community is doing. And it's DEFINITELY not what the standard library is designed for.
As someone who haven't done any systems programming after university: wait, what?
I was under impression that this is exactly what people where doing with Rust.(system apps, even linux kernel, no?)
If not - what do they (most if the community) are doing with Rust?
Some people definitely do systems programming in, but it's a minority. The std library is not set up for it at all, you need something like rustix, but even that results in very unidiomatic ("unsafe") rust code.
In Zig it's all in the std library by default. Because it's a systems programming language, first and foremost.
As far as I can tell, it contains many, many features that are irrelevant outside of systems programming scenarios with highly particular needs.
In zig it's std.posix.mmap.
And I don't think that including every feature of every possible OS is a sensible position to have for a standard library. Or would you argue that it should also include things like `CreateWindowExW()`?
If all you use is the Rust standard library, you can be reasonably sure that your program works on all platforms, and memory mapping is something that is highly platform specific, even among POSIX-likes. I would not like to attempt designing a singular cross-platform API for it that has to be maintained in perpetuity.
(There are a few OS-specific APIs in the Rust standard library, mostly because they are required anyway for things like I/O and process management. But the limit has to be set somewhere.)
In Zig it's exactly the same except that they decided to provide the POSIX platform specific APIs, which if you're using a POSIX system is I guess useful and otherwise it's dead weight.
It's a choice. I don't think it's a good choice but it's a choice.
extern "C" mmap(addr:*mut c_void, length:c_size_t, prot:c_int, flags:c_int, fd:c_int, offset:c_ssize_t) -> *mut c_void;
Zig has become my go-to for projects where I would previously have reached for C, largely because Zig has such good compatibility with other C projects.
Rust, on the other hand, is a completely different beast. It is very different from C, and it is far more complicated. That makes it harder to justify using, whereas Zig is a very easy choice as an alternative to using C itself.
Probably does something cool for all that crazy though?
This is a generic method in the middle of some database DSL code that does a bunch of SQL operations on a type safe manner. Code like this takes "SELECT ?+* FROM ?+* WHERE ? ORDER BY ?+* LIMIT ? OFFSET ?", specifically the limit and offset part, and returns a type that will always map to the database column. If the query is selecting a count of how many Foo each Baz references, this will map to a paginated Foo to Baz count type.
The alternative is to manually write this stuff out in SQL, then manually cast the right types into the basic primitives, which is what a language like Zig probably does.
You'll find similar (though sometimes less type-safe) complex code in just about any ORM/DSL, whether it's written in Java or PHP.
I don't think you can accomplish this in C without some kind of recursive macro parser generating either structs or maybe function pointers on the fly. It'd be hell to make that stuff not leak or double free memory, though.
But even so, I can't for the life of me write C code that's as safe as Rust. There are just too many ways to make subtle little mistakes here and there, incrementing a typed pointer by a sizeof by mistake thinking it's a uintptr_t, losing track of ownership and getting a use-after-free, messing up atomic access, mutex deadlocks oh my...
And that's with ALL warnings enabled in CLANG. It's even worse with the default warnings.
If you are writing more complicated or "interesting" programs, then I agree, C doesn't give you a good set of tools. But if all you are writing is small libraries or utility programs, C is just fine. In these cases, Rust feels like pulling out a sniper rifle to shoot a target a meter in front of your face (i.e., overkill).
If you are writing complex, large, or very mission-critical programs, then Rust is great to have as a tool as well. But we don't have to take such a black and white view to think that Rust is always the best tool for the job. Or C or Zig or whatever languages for that matter.
But the original claim was that "C code crashes all the time" which is blatantly wrong.
This is just plain not true. The fact of the matter is that if you stick to good practices, writing C code that does not crash under regular usage is not at all difficult. The problem is that writing C code that is guaranteed not to contain memory issues is extremely hard. But writing programs that don't crash "all the time" is not at all difficult.
This is why I can understand the push to use memory-safe languages to avoid memory safety issues in critical or widely-relied-upon programs. But the main reason you are using it in these cases is to avoid memory safety bugs, not to help you write programs that don't "crash all the time". And even then, the main incentive is that memory safety bugs can be quite severe bugs in areas like networking code, where they might become a significant security risk. The crashes aren't even the main concern here.
On using C from the creators of curl:
> We count about 40% of our security vulnerabilities to date to have been the direct result of us using C instead of a memory-safe language alternative. (https://daniel.haxx.se/blog/2025/04/07/writing-c-for-curl)
And, like what I was referring to, it is even more difficult to prove the absence of memory safety issues entirely. We have only managed to actually prove it for a handful of non-trivial programs, like the seL4 microkernel.
Now, this is all an entirely different question to whether the extra work required to write your program in Rust to avoid memory safety problems is actually worthwhile. If you are writing a program with no networking code, no privilege escalation, and no confidential information, the answer is probably that it doesn't matter. If a game has a buffer overflow in some weird edge case, it doesn't really matter. If your data processing code has a problem, you can probably spot that and it probably won't lead to any security concerns. If your file system scanner runs into a memory safety issue, it probably won't negatively impact you that much.
But if your crypto library has memory safety problems, you are in for a bad time.
When a project grows in complexity, age, number of contributors, poor review culture, complex requirements, or any of 101 different things that can make a project hard, then it is hard to avoid memory safety bugs. This covers like 99% of non-trivial projects.
If your software projects are very simple, then it is easy to write simple C programs that do the job well. This is what I said before. If your project is simple, boring, and straight-forward, using C is just fine. It is complex projects where the use of C can become a problem (and complexity can sneak up on you from an unbelievable number of sources).
Do you consider Rust enums 'proper sum types'? If yes what are Zig's tagged unions missing?
E.g.:
const Variant = union(enum) {
int: i32,
boolean: bool,
none,
fn truthy(self: Variant) bool {
return switch (self) {
Variant.int => |x_int| x_int != 0,
Variant.boolean => |x_bool| x_bool,
Variant.none => false,
};
}
};As for why you would choose it over C, because C has too many problems for even the C lovers to ignore. Zig fixes a tiny amount of them, just enough to pretend it's not problematic, but not enough to be useful in any non-hobby capacity. Which is fine, very few languages do achieve non-hobby status after all.